Optimisation method with node selection and centroid algorithm in underwater received signal strength localisation

Target localisation is one of the key technologies and a performance metric in underwater acoustic sensor networks. A number of range-free and range-based localisation algorithms are proposed. Among them a range-based localisation algorithm: weighted centroid localisation (WCL) is widely used due to the relatively simple implementation and better localisation. The accuracy of WCL, however, is reduced when the sensor nodes are farther from the centre or inappropriate nodes are selected to perform the localisation. This study proposes a novel and an efficient node optimisation algorithm based on received signal strength and weighted centroid. The proposed algorithm consists of three stages: (a) to estimate a rough region of the target node using bearing line and azimuth; (b) to calculate the optimal localisation region by the optimal localisation region function, and then (c) compare the overlap area of the rough region of the target node and the optimal localisation region to select the optimal localisation node. Simulation experiments show that the proposed optimal localisation region function can accurately represent the optimal localisation region of the geometrical figure, composed of different nodes. The study compares the performance of our proposed algorithm with the typical WCL algorithm, and clearly shows significant improvement in the localisation accuracy.